Claude Adds Excel Integration And Seven Connectors

Anthropic, delving deeper into financial workflows, introduces direct support for Microsoft Excel, along with seven new real-time market connectors and a portfolio of finance-specific Agent Skills.

The move places Claude inside the minds of hands-on assistants for analysts, bankers, and CFO teams who operate their lives within spreadsheets and make decisions with time-sensitive data.

Excel: Claude’s New 5th Professional Room

The new Excel integration moves Claude into a sidebar within an open spreadsheet, enabling users to query tables, sanitize messy data, and generate fresh sheets without ruining carefully crafted models.

Think: write out a variance table in plain English, standardize the formats for columns, or spin up a comps sheet from a series of tickers while keeping your formulas and structure intact.

For finance teams, that matters. Excel is still the sector’s OS (operating system) for models, reporting, and ad hoc analysis. A copilot AI in the environment where work actually occurs cuts down on tool switching and reduces friction. And in early trials, companies often find that “explain and transform” tasks — data normalization, narrative drafts for management decks, quick ratio checks — are the first to rack up measurable time savings.

Claude for Excel is a preview feature, and available in the Max, Enterprise, and Teams paid tiers, which underscores that Anthropic now packages its platform enterprise-first. “Our stack increasingly aligns with Microsoft’s productivity stack.”

Live Market Action through Seven Connectors

Anthropic also debuted seven connectors, bringing updates on real-time information into Claude conversations. Aiera provides earnings call transcripts and investor event materials; MT Newswires offers market-moving news; Moody’s serves up credit research; and LSEG delivers market and reference data coverage. For private markets and deal workflows, Chronograph provides portfolio monitoring data, Egnyte links secure content from data rooms, and Third Bridge expert insights are available on Aiera.

The thread here is speed to insight. Rather than toggling between terminals, a user might issue the demand: “Summarize today’s updates on our coverage universe, flag any rating changes from Moody’s, and refresh the comps in my Excel file with LSEG pricing.” It’s that hybrid query style — source-inflected, action-centric — where AI assistants begin to feel more like junior analysts than chatbots.

Anthropic claims that the new connectors are open to all Claude users, extending beyond the premium Excel preview.

Six Finance Skills For Analyst-Grade Work

As an additional aid to the connectors, Anthropic released six finance-focused Agent Skills — prebuilt instruction sets that seek to standardize how Claude approaches routine tasks.

LineUp includes earnings analysis, comparable company analysis, discounted cash flow modeling (including the ability to set different scenarios), support for data room document processing, company profiles for pitchbooks and lists of buyers, and initiation-style performance reports.

In practice, Skills function as reusable playbooks. A research associate might dump a transcript into Excel, run the earnings analysis Skill, and get back an executive-summary memo, a KPI trend table, and follow-up questions for management — all written in your house style and formatted according to disclosure rules. Skills are now available in a research preview for Max, Enterprise, and Teams users.

Why This Is Important for Finance Leaders

Generative AI is going from small talk to domain insight. McKinsey estimates the value of generative AI could total $200–$340B a year for banking, mostly in customer operations, marketing, risk, and software engineering. Now, front-office research and deal teams are seeing the same kinds of gains as more assistants leave their data-entry corners and get around to real data and spreadsheets — where decisions can actually happen.

There is evidence that modern models are capable of carrying more complex delegations. Recent academic research indicates that the highest-end systems have passed the notoriously difficult CFA Level III exam, where typical pass rates are in the 40–50% range, according to the CFA Institute. It doesn’t turn AI into a portfolio manager, but it does suggest an increased ability to synthesize long-form financial materials and deploy structured reasoning.

Governance, Controls And The Reality Check

Enterprise buyers will continue to pick apart governance — source provenance, audit trails, retention controls, and the treatment of material nonpublic information. The Egnyte connector indicates a way forward, however: preserving sensitive files within governed storage while making AI-supervised review possible. Teams should combine new workflows with human sign-off and model validation, particularly when it comes to valuation outputs and forward-looking statements.

How It Stacks Up In The Crowded AI Tools Market

Claude’s sidebar in Excel and live data connectors reflect a larger trend in the market. Microsoft has used partner models with Copilot and Office, while Google, on its Finance experience, included real-time market signals. The distinguishing feature of Anthropic might then be the strong emphasis it places on controllable Skills and the attention to high-signal, finance-native sources like Aiera, Moody’s, and LSEG.

Bottom Line For Teams Considering Claude In Excel

If your analysts’ world revolves around Excel and your edge is based on speed to synthesis, this update could be worth piloting. Begin with highly constrained use cases — earnings summaries, comps refresh, credit update briefs, etc. — and measure cycle-time reduction and error rates. With mindful governance, such connectors and Skills can transform Claude from a generic chatbot into a useful junior teammate embedded with the tools that finance already trusts.